Prosper Loan Analysis by Gregory Winkler
This report explores approximately 114,000 Propser Loan Data for loans made from November 2005 to March 2014. Prosper is a marketplace lending platform that allows for peer to peer loans.
Univariate Plots Section
## Classes 'tbl_df', 'tbl' and 'data.frame': 113937 obs. of 83 variables:
## $ ListingKey : chr "1021339766868145413AB3B" "10273602499503308B223C1" "0EE9337825851032864889A" ...
## $ ListingNumber : int 193129 1209647 81716 658116 909464 1074836 750899 768193 ...
## $ ListingCreationDate : POSIXct, format: "2007-08-26 19:09:29" "2014-02-27 08:28:07" ...
## $ CreditGrade : chr "C" NA "HR" ...
## $ Term : Factor w/ 3 levels "12","36","60": 2 2 2 2 2 3 2 2 ...
## $ LoanStatus : Factor w/ 12 levels "Cancelled","Chargedoff",..: 3 4 3 4 4 4 4 4 ...
## $ ClosedDate : POSIXct, format: "2009-08-14" NA ...
## $ BorrowerAPR : num 0.165 0.12 0.283 0.125 ...
## $ BorrowerRate : num 0.158 0.092 0.275 0.0974 ...
## $ LenderYield : num 0.138 0.082 0.24 0.0874 ...
## $ EstimatedEffectiveYield : num NA 0.0796 NA 0.0849 ...
## $ EstimatedLoss : num NA 0.0249 NA 0.0249 ...
## $ EstimatedReturn : num NA 0.0547 NA 0.06 ...
## $ ProsperRating (numeric) : int NA 6 NA 6 3 5 2 4 ...
## $ ProsperRating (Alpha) : chr NA "A" NA ...
## $ ProsperScore : num NA 7 NA 9 4 10 2 4 ...
## $ ListingCategory (numeric) : int 0 2 0 16 2 1 1 2 ...
## $ BorrowerState : chr "CO" "CO" "GA" ...
## $ Occupation : Factor w/ 67 levels "Accountant/CPA",..: 36 42 36 51 20 42 49 28 ...
## $ EmploymentStatus : Factor w/ 8 levels "Employed","Full-time",..: 8 1 3 1 1 1 1 1 ...
## $ EmploymentStatusDuration : int 2 44 NA 113 44 82 172 103 ...
## $ IsBorrowerHomeowner : Factor w/ 2 levels "False","True": 2 1 1 2 2 2 1 1 ...
## $ CurrentlyInGroup : Factor w/ 2 levels "False","True": 2 1 2 1 1 1 1 1 ...
## $ GroupKey : chr NA NA "783C3371218786870A73D20" ...
## $ DateCreditPulled : POSIXct, format: "2007-08-26 18:41:46" "2014-02-27 08:28:14" ...
## $ CreditScoreRangeLower : int 640 680 480 800 680 740 680 700 ...
## $ CreditScoreRangeUpper : int 659 699 499 819 699 759 699 719 ...
## $ FirstRecordedCreditLine : POSIXct, format: "2001-10-11" "1996-03-18" ...
## $ CurrentCreditLines : int 5 14 NA 5 19 21 10 6 ...
## $ OpenCreditLines : int 4 14 NA 5 19 17 7 6 ...
## $ TotalCreditLinespast7years : int 12 29 3 29 49 49 20 10 ...
## $ OpenRevolvingAccounts : int 1 13 0 7 6 13 6 5 ...
## $ OpenRevolvingMonthlyPayment : num 24 389 0 115 220 1410 214 101 ...
## $ InquiriesLast6Months : int 3 3 0 0 1 0 0 3 ...
## $ TotalInquiries : num 3 5 1 1 9 2 0 16 ...
## $ CurrentDelinquencies : int 2 0 1 4 0 0 0 0 ...
## $ AmountDelinquent : num 472 0 NA 10056 ...
## $ DelinquenciesLast7Years : int 4 0 0 14 0 0 0 0 ...
## $ PublicRecordsLast10Years : int 0 1 0 0 0 0 0 1 ...
## $ PublicRecordsLast12Months : int 0 0 NA 0 0 0 0 0 ...
## $ RevolvingCreditBalance : num 0 3989 NA 1444 ...
## $ BankcardUtilization : num 0 0.21 NA 0.04 0.81 0.39 0.72 0.13 ...
## $ AvailableBankcardCredit : num 1500 10266 NA 30754 ...
## $ TotalTrades : num 11 29 NA 26 39 47 16 10 ...
## $ TradesNeverDelinquent (percentage) : num 0.81 1 NA 0.76 0.95 1 0.68 0.8 ...
## $ TradesOpenedLast6Months : num 0 2 NA 0 2 0 0 0 ...
## $ DebtToIncomeRatio : num 0.17 0.18 0.06 0.15 0.26 0.36 0.27 0.24 ...
## $ IncomeRange : Ord.factor w/ 8 levels "Not displayed"<..: 5 6 1 5 8 8 5 5 ...
## $ IncomeVerifiable : Factor w/ 2 levels "False","True": 2 2 2 2 2 2 2 2 ...
## $ StatedMonthlyIncome : num 3083 6125 2083 2875 ...
## $ LoanKey : chr "E33A3400205839220442E84" "9E3B37071505919926B1D82" "6954337960046817851BCB2" ...
## $ TotalProsperLoans : int NA NA NA NA 1 NA NA NA ...
## $ TotalProsperPaymentsBilled : int NA NA NA NA 11 NA NA NA ...
## $ OnTimeProsperPayments : int NA NA NA NA 11 NA NA NA ...
## $ ProsperPaymentsLessThanOneMonthLate: int NA NA NA NA 0 NA NA NA ...
## $ ProsperPaymentsOneMonthPlusLate : int NA NA NA NA 0 NA NA NA ...
## $ ProsperPrincipalBorrowed : num NA NA NA NA 11000 NA NA NA ...
## $ ProsperPrincipalOutstanding : num NA NA NA NA ...
## $ ScorexChangeAtTimeOfListing : int NA NA NA NA NA NA NA NA ...
## $ LoanCurrentDaysDelinquent : int 0 0 0 0 0 0 0 0 ...
## $ LoanFirstDefaultedCycleNumber : int NA NA NA NA NA NA NA NA ...
## $ LoanMonthsSinceOrigination : int 78 0 86 16 6 3 11 10 ...
## $ LoanNumber : int 19141 134815 6466 77296 102670 123257 88353 90051 ...
## $ LoanOriginalAmount : int 9425 10000 3001 10000 15000 15000 3000 10000 ...
## $ LoanOriginationDate : POSIXct, format: "2007-09-12" "2014-03-03" ...
## $ LoanOriginationQuarter : chr "Q3 2007" "Q1 2014" "Q1 2007" ...
## $ MemberKey : chr "1F3E3376408759268057EDA" "1D13370546739025387B2F4" "5F7033715035555618FA612" ...
## $ MonthlyLoanPayment : num 330 319 123 321 ...
## $ LP_CustomerPayments : num 11396 0 4187 5143 ...
## $ LP_CustomerPrincipalPayments : num 9425 0 3001 4091 ...
## $ LP_InterestandFees : num 1971 0 1186 1052 ...
## $ LP_ServiceFees : num -133.2 0 -24.2 -108 ...
## $ LP_CollectionFees : num 0 0 0 0 0 0 0 0 ...
## $ LP_GrossPrincipalLoss : num 0 0 0 0 0 0 0 0 ...
## $ LP_NetPrincipalLoss : num 0 0 0 0 0 0 0 0 ...
## $ LP_NonPrincipalRecoverypayments : num 0 0 0 0 0 0 0 0 ...
## $ PercentFunded : num 1 1 1 1 1 1 1 1 ...
## $ Recommendations : int 0 0 0 0 0 0 0 0 ...
## $ InvestmentFromFriendsCount : int 0 0 0 0 0 0 0 0 ...
## $ InvestmentFromFriendsAmount : num 0 0 0 0 0 0 0 0 ...
## $ Investors : int 258 1 41 158 20 1 1 1 ...
## $ LoanOriginationDate.bucket : Factor w/ 10 levels "(2005,2006]",..: 2 9 2 7 8 8 8 8 ...
## $ BankcardUtilization.bucket : Factor w/ 5 levels "(0,0.25]","(0.25,0.5]",..: NA 1 NA 1 4 2 3 1 ...## ListingKey ListingNumber ListingCreationDate
## Length:113937 Min. : 4 Min. :2005-11-09 20:44:28
## Class :character 1st Qu.: 400919 1st Qu.:2008-09-19 10:02:14
## Mode :character Median : 600554 Median :2012-06-16 12:37:19
## Mean : 627886 Mean :2011-07-09 08:07:23
## 3rd Qu.: 892634 3rd Qu.:2013-09-09 19:40:48
## Max. :1255725 Max. :2014-03-10 12:20:53
##
## CreditGrade Term LoanStatus
## Length:113937 12: 1614 Current :56576
## Class :character 36:87778 Completed :38074
## Mode :character 60:24545 Chargedoff :11992
## Defaulted : 5018
## Past Due (1-15 days) : 806
## Past Due (31-60 days): 363
## (Other) : 1108
## ClosedDate BorrowerAPR BorrowerRate
## Min. :2005-11-25 00:00:00 Min. :0.00653 Min. :0.0000
## 1st Qu.:2009-07-14 00:00:00 1st Qu.:0.15629 1st Qu.:0.1340
## Median :2011-04-05 00:00:00 Median :0.20976 Median :0.1840
## Mean :2011-03-07 20:21:21 Mean :0.21883 Mean :0.1928
## 3rd Qu.:2013-01-30 00:00:00 3rd Qu.:0.28381 3rd Qu.:0.2500
## Max. :2014-03-10 00:00:00 Max. :0.51229 Max. :0.4975
## NA's :58848 NA's :25
## LenderYield EstimatedEffectiveYield EstimatedLoss
## Min. :-0.0100 Min. :-0.183 Min. :0.005
## 1st Qu.: 0.1242 1st Qu.: 0.116 1st Qu.:0.042
## Median : 0.1730 Median : 0.162 Median :0.072
## Mean : 0.1827 Mean : 0.169 Mean :0.080
## 3rd Qu.: 0.2400 3rd Qu.: 0.224 3rd Qu.:0.112
## Max. : 0.4925 Max. : 0.320 Max. :0.366
## NA's :29084 NA's :29084
## EstimatedReturn ProsperRating (numeric) ProsperRating (Alpha)
## Min. :-0.183 Min. :1.000 Length:113937
## 1st Qu.: 0.074 1st Qu.:3.000 Class :character
## Median : 0.092 Median :4.000 Mode :character
## Mean : 0.096 Mean :4.072
## 3rd Qu.: 0.117 3rd Qu.:5.000
## Max. : 0.284 Max. :7.000
## NA's :29084 NA's :29084
## ProsperScore ListingCategory (numeric) BorrowerState
## Min. : 1.00 Min. : 0.000 Length:113937
## 1st Qu.: 4.00 1st Qu.: 1.000 Class :character
## Median : 6.00 Median : 1.000 Mode :character
## Mean : 5.95 Mean : 2.774
## 3rd Qu.: 8.00 3rd Qu.: 3.000
## Max. :11.00 Max. :20.000
## NA's :29084
## Occupation EmploymentStatus
## Other :28617 Employed :67322
## Professional :13628 Full-time :26355
## Computer Programmer: 4478 Self-employed: 6134
## Executive : 4311 Not available: 5347
## Teacher : 3759 Other : 3806
## (Other) :55556 (Other) : 2718
## NA's : 3588 NA's : 2255
## EmploymentStatusDuration IsBorrowerHomeowner CurrentlyInGroup
## Min. : 0.00 False:56459 False:101218
## 1st Qu.: 26.00 True :57478 True : 12719
## Median : 67.00
## Mean : 96.07
## 3rd Qu.:137.00
## Max. :755.00
## NA's :7625
## GroupKey DateCreditPulled CreditScoreRangeLower
## Length:113937 Min. :2005-11-09 00:30:04 Min. : 0.0
## Class :character 1st Qu.:2008-09-16 22:25:27 1st Qu.:660.0
## Mode :character Median :2012-06-17 07:52:34 Median :680.0
## Mean :2011-07-09 15:28:40 Mean :685.6
## 3rd Qu.:2013-09-11 14:30:24 3rd Qu.:720.0
## Max. :2014-03-10 12:20:56 Max. :880.0
## NA's :591
## CreditScoreRangeUpper FirstRecordedCreditLine CurrentCreditLines
## Min. : 19.0 Min. :1947-08-24 00:00:00 Min. : 0.00
## 1st Qu.:679.0 1st Qu.:1990-06-01 00:00:00 1st Qu.: 7.00
## Median :699.0 Median :1995-11-01 00:00:00 Median :10.00
## Mean :704.6 Mean :1994-11-17 07:00:07 Mean :10.32
## 3rd Qu.:739.0 3rd Qu.:2000-03-14 00:00:00 3rd Qu.:13.00
## Max. :899.0 Max. :2012-12-22 00:00:00 Max. :59.00
## NA's :591 NA's :697 NA's :7604
## OpenCreditLines TotalCreditLinespast7years OpenRevolvingAccounts
## Min. : 0.00 Min. : 2.00 Min. : 0.00
## 1st Qu.: 6.00 1st Qu.: 17.00 1st Qu.: 4.00
## Median : 9.00 Median : 25.00 Median : 6.00
## Mean : 9.26 Mean : 26.75 Mean : 6.97
## 3rd Qu.:12.00 3rd Qu.: 35.00 3rd Qu.: 9.00
## Max. :54.00 Max. :136.00 Max. :51.00
## NA's :7604 NA's :697
## OpenRevolvingMonthlyPayment InquiriesLast6Months TotalInquiries
## Min. : 0.0 Min. : 0.000 Min. : 0.000
## 1st Qu.: 114.0 1st Qu.: 0.000 1st Qu.: 2.000
## Median : 271.0 Median : 1.000 Median : 4.000
## Mean : 398.3 Mean : 1.435 Mean : 5.584
## 3rd Qu.: 525.0 3rd Qu.: 2.000 3rd Qu.: 7.000
## Max. :14985.0 Max. :105.000 Max. :379.000
## NA's :697 NA's :1159
## CurrentDelinquencies AmountDelinquent DelinquenciesLast7Years
## Min. : 0.0000 Min. : 0.0 Min. : 0.000
## 1st Qu.: 0.0000 1st Qu.: 0.0 1st Qu.: 0.000
## Median : 0.0000 Median : 0.0 Median : 0.000
## Mean : 0.5921 Mean : 984.5 Mean : 4.155
## 3rd Qu.: 0.0000 3rd Qu.: 0.0 3rd Qu.: 3.000
## Max. :83.0000 Max. :463881.0 Max. :99.000
## NA's :697 NA's :7622 NA's :990
## PublicRecordsLast10Years PublicRecordsLast12Months RevolvingCreditBalance
## Min. : 0.0000 Min. : 0.000 Min. : 0
## 1st Qu.: 0.0000 1st Qu.: 0.000 1st Qu.: 3121
## Median : 0.0000 Median : 0.000 Median : 8549
## Mean : 0.3126 Mean : 0.015 Mean : 17599
## 3rd Qu.: 0.0000 3rd Qu.: 0.000 3rd Qu.: 19521
## Max. :38.0000 Max. :20.000 Max. :1435667
## NA's :697 NA's :7604 NA's :7604
## BankcardUtilization AvailableBankcardCredit TotalTrades
## Min. :0.000 Min. : 0 Min. : 0.00
## 1st Qu.:0.310 1st Qu.: 880 1st Qu.: 15.00
## Median :0.600 Median : 4100 Median : 22.00
## Mean :0.561 Mean : 11210 Mean : 23.23
## 3rd Qu.:0.840 3rd Qu.: 13180 3rd Qu.: 30.00
## Max. :5.950 Max. :646285 Max. :126.00
## NA's :7604 NA's :7544 NA's :7544
## TradesNeverDelinquent (percentage) TradesOpenedLast6Months
## Min. :0.000 Min. : 0.000
## 1st Qu.:0.820 1st Qu.: 0.000
## Median :0.940 Median : 0.000
## Mean :0.886 Mean : 0.802
## 3rd Qu.:1.000 3rd Qu.: 1.000
## Max. :1.000 Max. :20.000
## NA's :7544 NA's :7544
## DebtToIncomeRatio IncomeRange IncomeVerifiable
## Min. : 0.000 $25,000-49,999:32192 False: 8669
## 1st Qu.: 0.140 $50,000-74,999:31050 True :105268
## Median : 0.220 $100,000+ :17337
## Mean : 0.276 $75,000-99,999:16916
## 3rd Qu.: 0.320 Not displayed : 7741
## Max. :10.010 $1-24,999 : 7274
## NA's :8554 (Other) : 1427
## StatedMonthlyIncome LoanKey TotalProsperLoans
## Min. : 0 Length:113937 Min. :0.00
## 1st Qu.: 3200 Class :character 1st Qu.:1.00
## Median : 4667 Mode :character Median :1.00
## Mean : 5608 Mean :1.42
## 3rd Qu.: 6825 3rd Qu.:2.00
## Max. :1750003 Max. :8.00
## NA's :91852
## TotalProsperPaymentsBilled OnTimeProsperPayments
## Min. : 0.00 Min. : 0.00
## 1st Qu.: 9.00 1st Qu.: 9.00
## Median : 16.00 Median : 15.00
## Mean : 22.93 Mean : 22.27
## 3rd Qu.: 33.00 3rd Qu.: 32.00
## Max. :141.00 Max. :141.00
## NA's :91852 NA's :91852
## ProsperPaymentsLessThanOneMonthLate ProsperPaymentsOneMonthPlusLate
## Min. : 0.00 Min. : 0.00
## 1st Qu.: 0.00 1st Qu.: 0.00
## Median : 0.00 Median : 0.00
## Mean : 0.61 Mean : 0.05
## 3rd Qu.: 0.00 3rd Qu.: 0.00
## Max. :42.00 Max. :21.00
## NA's :91852 NA's :91852
## ProsperPrincipalBorrowed ProsperPrincipalOutstanding
## Min. : 0 Min. : 0
## 1st Qu.: 3500 1st Qu.: 0
## Median : 6000 Median : 1627
## Mean : 8472 Mean : 2930
## 3rd Qu.:11000 3rd Qu.: 4127
## Max. :72499 Max. :23451
## NA's :91852 NA's :91852
## ScorexChangeAtTimeOfListing LoanCurrentDaysDelinquent
## Min. :-209.00 Min. : 0.0
## 1st Qu.: -35.00 1st Qu.: 0.0
## Median : -3.00 Median : 0.0
## Mean : -3.22 Mean : 152.8
## 3rd Qu.: 25.00 3rd Qu.: 0.0
## Max. : 286.00 Max. :2704.0
## NA's :95009
## LoanFirstDefaultedCycleNumber LoanMonthsSinceOrigination LoanNumber
## Min. : 0.00 Min. : 0.0 Min. : 1
## 1st Qu.: 9.00 1st Qu.: 6.0 1st Qu.: 37332
## Median :14.00 Median : 21.0 Median : 68599
## Mean :16.27 Mean : 31.9 Mean : 69444
## 3rd Qu.:22.00 3rd Qu.: 65.0 3rd Qu.:101901
## Max. :44.00 Max. :100.0 Max. :136486
## NA's :96985
## LoanOriginalAmount LoanOriginationDate LoanOriginationQuarter
## Min. : 1000 Min. :2005-11-15 00:00:00 Length:113937
## 1st Qu.: 4000 1st Qu.:2008-10-02 00:00:00 Class :character
## Median : 6500 Median :2012-06-26 00:00:00 Mode :character
## Mean : 8337 Mean :2011-07-21 03:18:19
## 3rd Qu.:12000 3rd Qu.:2013-09-18 00:00:00
## Max. :35000 Max. :2014-03-12 00:00:00
##
## MemberKey MonthlyLoanPayment LP_CustomerPayments
## Length:113937 Min. : 0.0 Min. : -2.35
## Class :character 1st Qu.: 131.6 1st Qu.: 1005.76
## Mode :character Median : 217.7 Median : 2583.83
## Mean : 272.5 Mean : 4183.08
## 3rd Qu.: 371.6 3rd Qu.: 5548.40
## Max. :2251.5 Max. :40702.39
##
## LP_CustomerPrincipalPayments LP_InterestandFees LP_ServiceFees
## Min. : 0.0 Min. : -2.35 Min. :-664.87
## 1st Qu.: 500.9 1st Qu.: 274.87 1st Qu.: -73.18
## Median : 1587.5 Median : 700.84 Median : -34.44
## Mean : 3105.5 Mean : 1077.54 Mean : -54.73
## 3rd Qu.: 4000.0 3rd Qu.: 1458.54 3rd Qu.: -13.92
## Max. :35000.0 Max. :15617.03 Max. : 32.06
##
## LP_CollectionFees LP_GrossPrincipalLoss LP_NetPrincipalLoss
## Min. :-9274.75 Min. : -94.2 Min. : -954.5
## 1st Qu.: 0.00 1st Qu.: 0.0 1st Qu.: 0.0
## Median : 0.00 Median : 0.0 Median : 0.0
## Mean : -14.24 Mean : 700.4 Mean : 681.4
## 3rd Qu.: 0.00 3rd Qu.: 0.0 3rd Qu.: 0.0
## Max. : 0.00 Max. :25000.0 Max. :25000.0
##
## LP_NonPrincipalRecoverypayments PercentFunded Recommendations
## Min. : 0.00 Min. :0.7000 Min. : 0.00000
## 1st Qu.: 0.00 1st Qu.:1.0000 1st Qu.: 0.00000
## Median : 0.00 Median :1.0000 Median : 0.00000
## Mean : 25.14 Mean :0.9986 Mean : 0.04803
## 3rd Qu.: 0.00 3rd Qu.:1.0000 3rd Qu.: 0.00000
## Max. :21117.90 Max. :1.0125 Max. :39.00000
##
## InvestmentFromFriendsCount InvestmentFromFriendsAmount Investors
## Min. : 0.00000 Min. : 0.00 Min. : 1.00
## 1st Qu.: 0.00000 1st Qu.: 0.00 1st Qu.: 2.00
## Median : 0.00000 Median : 0.00 Median : 44.00
## Mean : 0.02346 Mean : 16.55 Mean : 80.48
## 3rd Qu.: 0.00000 3rd Qu.: 0.00 3rd Qu.: 115.00
## Max. :33.00000 Max. :25000.00 Max. :1189.00
##
## LoanOriginationDate.bucket BankcardUtilization.bucket
## (2012,2013]:34345 (0,0.25] :15842
## (2011,2012]:19553 (0.25,0.5]:20866
## (2013,2014]:12172 (0.5,0.75]:26578
## (2007,2008]:11552 (0.75,1] :34521
## (2006,2007]:11460 (1,5] : 1742
## (Other) :24833 NA's :14388
## NA's : 22
## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.0 660.0 680.0 685.6 720.0 880.0 591The credit scores have a median of 680.
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.0000 0.1340 0.1840 0.1928 0.2500 0.4975The borrower’s interest rates have a normal distribution with the exception of a significant number of loans around 0.33 (33%).
The largest group of borrowers have incomes in the $25,000 to $74,999 Income Ranges.
## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.000 0.140 0.220 0.276 0.320 10.010 8554The Debt-to-Income Ratio for most borrowers is reasonable with a mean around 0.27. However, there is a sizable number of borrowers over 10. Such a high number would be difficult to ever pay back. Taking a quick look at the IncomeRanges and whether or not that income is verifiable shows that these borrowers are predominantly low income and possibly may not be disclosing their true income.
## Source: local data frame [8 x 3]
## Groups: IncomeRange [?]
##
## IncomeRange IncomeVerifiable n
## <ord> <fctr> <int>
## 1 Not displayed False 49
## 2 Not displayed True 10
## 3 Not employed False 21
## 4 Not employed True 3
## 5 $1-24,999 False 115
## 6 $1-24,999 True 72
## 7 $50,000-74,999 True 1
## 8 $100,000+ True 1
Excluding the highest 1% of debt-to-income ratios, we see that the majority of borrowers are centered around 0.2.
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1000 4000 6500 8337 12000 35000Median loan amount is $6,500, but range from as low as $1,000 up to $35,000. Amounts are most frequent in $5,000 intervals.
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.0 131.6 217.7 272.5 371.6 2252.0Monthly loan payments are for the most part below $500.
Dividing the monthly loan payment by the borrower’s stated monthly income, we see that most borrowers have a monthly payment that is less than 5% of their income. However there is a significant right skew to this graph. I created this ratio because it stands to reason that borrowers paying a high amount of their monthly income will struggle to repay their loan.
## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0.000 0.310 0.600 0.561 0.840 5.950 7604Bankcard Utilization shows a large number of borrowers that have either used none of their available credit or all of it.
## # A tibble: 3 × 2
## `loanData$DelinquenciesLast7Years > 0` n
## <lgl> <int>
## 1 FALSE 76439
## 2 TRUE 36508
## 3 NA 990Of the more than 36,000 borrowers that had delinquent accounts within the last 7 years on their credit report, the majority had less than 5.
## # A tibble: 3 × 2
## `loanData$InquiriesLast6Months > 0` n
## <lgl> <int>
## 1 FALSE 50005
## 2 TRUE 63235
## 3 NA 697Many of the borrowers did not have any credit inquiries in the 6 months leading up to their loans.
Most of the loans were made in 2013 and 2014, the last 2 years of the data. Note the gap with no loans in late 2008 and first half of 2009.
The majority of of loans are short term at 36 months to maturity.
Debt Consolidation appears as the most frequent reason for borrowing.
The high charge-offs and defaulted loans in this dataset are worth taking a closer look at to see if there are any possible causes. Separately, the high amount of completed loans is expected given that the majority of loans are 36 months and the dataset spans just shy of 10 years.
California leads the way with the most borrowers, with some of the other more populated states trailing behind.
Univariate Analysis
Observations
There are 113,937 loan records in the dataset with 81 variables about the borrower and their loan. Some of the key variables are the borrower rate, credit score, debt-to-income ratio, bankcard utilization rate.
- Most of the borrowers have a credit score of approximately 680.
- Median borrower rate is 0.184, but exhibits a right skew with a spike around 0.33.
- Median debt-to-income ratio is 0.22.
- Median bankcard utilization is 0.60 with most borrowers at 0 or 1.
Further Exploration
I am interested in taking a closer look at the borrower’s credit history to see what relationships there may be with a borrower’s interest rate. Based on some of the plots we saw earlier, I would expect there to be a few variables that explain some of the rate differentials. Some variables that might contribute are the number of delinquencies, credit inquiries, debt-to-income ratio, and their bankcard utilization rate. Also, some of the qualitative variables will be interesting to look at to see if they contributed to any disparity. For example, does the borrower’s geography or purpose of loan have a noticeable impact.
Also, this dataset provides an interesting time frame in that it includes loans made at the peak of the economic cycle (2006-2007) through one of the worst financial collapses in recent history. I would expect defaults to be high in 2009 and 2010 as many people became unemployed.
The time period of this data should be an interesting factor as the financial collapse in 2008-2009 resulted in many people losing their jobs and therefore income. I would expect defaults to be high in 2008 through 2010 as many people became unemployed.
Data Adjustments
I made some adjustment to the dataset. Sepecifically, I created a new variable for the Listing Category to show what it is rather than a vague numerical value. Also, I created Credit Score Buckets to group credit scores for my analysis.
Bivariate Plots Section
A quick look at a few of the variable relationships shows some meaningful correlations with the borrower’s credit score. Also, the credit score appears to have a fairly strong negative correlation with the borrower’s rate as would be expected.
I quickly notice a couple of meaningful correlations for variables relative to the borrower’s credit score.
I took a subset of the data here to only include borrowers with a bankcard utilization below 1.5. We can see the correlation between the two variables well here as the range and median of bankcard utilization rises as credit scores decline.
Bankcard Utilization rises slightly with income.
Inquiries in the last 6 months do not appear as meaningful as I was expecting. Typically 1 inquiry on a credit report will not cause a drop in a credit score. However, there does appear to be some significance to being above or below a credit score of 700.
Higher current delinquencies tend to be associated with lower credit scores. The outliers for delinquencies begin to trend higher starting with borrowers below 700 credit scores.
Debt-to-income shows normal distribution across credit scores and appears to have no correlation.
Lower credit quality borrowers appear more likely to fall behind on their payments.
Loan defaults were much greater in the time period leading up to the financial crisis then afterwards. The lower credit score borrowers clearly were more likely to default prior to the crisis and again around 2011 and 2012.
Credit scores show some correlation with the borrower’s interest rate with an R2 of -0.46.
Loan purposes listed as Not Available have the greatest outliers. Cosmetic procedures and Household expenses have the highest rates on average.
As a percentage of the total loans in a state, the northeast region looks to be the least likely area to default, whereas the northwest and midwest show the greatest potential for default.
Bivariate Analysis
Observations
It was interesting to note some of the variables that had a correlation with the credit score. As expected, the bankcard utilization, current delinquencies and number of inquiries in the last 6 months all exhibited some correlation. However, I was a little surprised that the debt-to-income ratio showed no correlation. While income may not factor into the credit score, I would have expected an indirect relationship here where borrowers that have high debt-to-income are more likely to fall behind on payments and/or utilize a large portion of their available credit.
Near the beginning of the analysis it was noted that a spike in borrowing rates was observed around 33%. This appeared with some surprising results when graphed against credit scores. The data showed a negative correlation between credit scores and borrower rates, but for the subset of data with rates around 33% showed a high concentration corresponding to borrowers with credit scores above what might be expected. Based on the graph of the relationship, I would not have expected many loan rates to coincide with credit scores over 700.
The number of credit inquiries and current delinquencies both showed that a higher occurrence in either was more likely to be associated with borrowers having lower credit scores.
Defaults were more frequent among low credit score borrowers which was to be expected. However, the more interesting aspect of the graph was the impact that exogenious factors have on the repayment of a loan. While a credit score can quantify a borrower’s ability to repay their loan based on historical information, it is unable to anticipate future variants that can lead to unexpected results. This is evident in the high frequency of defaults preceeding the 2008-2009 financial crisis when many people became unemployed.
The rise in borrower rates in 2011 through 2012 was interesting to note beause the FHLB Boston 3 year fully amortizing rate, a benchmark rate used by commercial banks, stayed at very low levels during this period. Since credit scores are representative of certain characteristic traits, it means that either lenders were demanding a higher risk premium or borrowers were taking loans out for longer terms.
The strongest relationship in the data I explored was the Credit Score and the Borrower’s Rate with a R2 of -0.46. The Bankcard Utilization rate and the Credit Scores was not significantly different though with a R2 of -0.40. Also, Current Delinquencies and Credit Inquiries showed some correlation with Credit Scores.
Multivariate Plots Section
In general, a higher bankcard utilization rate corresponds with lower credit scores and higher interest rates.
Loans that are past due tend to be higher rates and borrowers that have low credit scores and high bankcardutilization.
Graphing only the loans with interest rates between 31% and 34% having a credit score greater than 700, we see that these loans were almost entirely originated in 2011 and 2012.
Graphing the median borrowing rates over time, we see that interest rates started rising in 2010 and remained elevated through most of 2012. Meanwhile the FHLB Boston 3 year amortizing rate, a lending benchmark rate, remained at very low levels.
There is a slight increase in the credit risk premium in 2011 that may explain some of the rise in borrower rates observed.
Separately we see a significant increase in loans to borrowers with 675 to 700 credit scores starting in 2011 and increasing at a significantly faster pace in 2013.
It appears that 12 and 60 month loans are new to Prosper starting around 2011.
There has been an increase in 60 month term loans for borrowers with credit scores ranging from 700 to 775. However, the concentration around the 33% interest rate looks to be 36 month loans.
The 60 month loans do not make a noticable increase until 2012.
For the selected range of credit scores, it does appear that bankcard utilization for higher income borrowers may be a contributing factor for the higher rates.
Whereas higher debt-to-income may be more likely a contributing factor for lower income borrowers.
Multivariate Analysis
Observations
The median interest rates of the credit score ranges over the relevent period was insightful to provide a possible reason for why the data did not necessarily fit as well with other variables as I might have expected. Considering many of these borrowers are high risk, I would not be surprised if the increase in overall borrower rates during 2011 was from lenders being more conservative and expecting a higher return.
While bankcard utilization varied greatly, in general it was noticeably higher for loans that had a low credit score and high interest rate. Also, it looks to be a contributing factor to loans that are currently past due. This was inline with my expectations as these borrowers were most likely already struggling with repaying their debt.
As interest rate declined in 2013 and 2014, borrowers took advantage of the 60 month term option in increasing numbers.
Interesting Interactions between Features
It was interesting to see that for the interest rates between 31% and 34% of higher credit score borrowers, bankcard utilization had a bigger factor for higher income borrowers. This was in contrast to debt-to-income that appeared to be more influential for lower income borrowers.
Final Plots and Summary
Plot One
Description One
The difference in number of defaulted loans before and after the financial crisis was very interesting. This shows how our models and research can only provide insight on what we can expect based on historical examples. Also, we can see that these borrowers are of lower quality than the overall dataset with a median credit score of 640 and borrowing rate of 22.7%.
Credit Scores
| Min. | 1st Qu. | Median | Mean | 3rd Qu. | Max. |
|---|---|---|---|---|---|
| 420.0 | 560.0 | 640.0 | 625.3 | 680.0 | 860.0 |
Borrower Rates
| Min. | 1st Qu. | Median | Mean | 3rd Qu. | Max. |
|---|---|---|---|---|---|
| 0.00% | 16.50% | 22.70% | 22.15% | 28.50% | 36.00% |
Plot Two
Description Two
The relationship between credit scores and borrower rates shows that there is a negative correlation (R2 of -0.46) This makes sense as we would expect that the reason for the borrower’s low credit score is that they would be more likely to not pay back their loan in full.
| Credit Score Bucket | Median Bankcard Utilization |
|---|---|
| (400,600] | 0.86 |
| (600,650] | 0.78 |
| (650,675] | 0.73 |
| (675,700] | 0.65 |
| (700,725] | 0.52 |
| (725,775] | 0.38 |
| (775,900] | 0.18 |
Plot Three
Description Three
The median interest rates by credit score over time provide a couple of useful pieces of information. The show that lower credit borrowers are likely to have higher interest rates at any given point in time. Additionally, they show that during time periods of higher risk, these low credit borrowers may see larger differentials in pricing from other borrowers.
| Credit Score Bucket | Median Rate |
|---|---|
| (400,600] | 25.00% |
| (600,650] | 23.20% |
| (650,675] | 20.99% |
| (675,700] | 18.25% |
| (700,725] | 15.20% |
| (725,775] | 13.50% |
| (775,900] | 9.99% |
Reflection
The Prosper dataset was interesting to explore as there was information about the borrower’s historical experience with debt and also their current loan and its repayment performance. It was relatively easy to work with the data as most of it was already in a workable format. However, there were a few variables that I found helpful to factor first to make them easier to work with.
It was interesting to bring in external interest rate data into the dataset. The Prosper data provided a lot of interesting information but there were a couple of times that I thought a stronger relationship should have existed but did not. Understanding the general interest rate market conditions helped to guide further exploration. This was most helpful with exploring the large spike in loan rates around 33%.
In the future, I think the analysis could be expanded to try to predict interest rates on new loans. I think there is enough information from the propser data that when combined with general market yield curves, it could come close to providing some prediction on where new loans would be given a set of borrower characteristics. However, Prosper’s business model likely makes this a little more difficult in that at any point in time there could be an imbalance of borrowers and lenders since the business is not universally known as a typical bank would be.
Additional data sources
- FHLB Boston Rates: http://www.fhlbboston.com/rates/historicalrates/index.jsp
- Bank of America Merrill Lynch OAS: https://fred.stlouisfed.org/series/BAMLH0A1HYBB